Precursors and Atmospheric Phenomena
Portraying earthquakes as events with explainable mechanical causes was part of the strategy in Thoughts on Earthquakes for promoting calm. Significantly, the book extends this approach to suggest that earthquakes might be predictable. The imbalance of yin and yang beneath the earth might manifest itself in a variety of precursor phenomena (zenchō). Although claims of earthquake precursors were not new in 1830, Thoughts on Earthquakes was the earliest widely read book to compile these alleged precursors with an eye to the future. We saw that Nishikawa Joken expressed skepticism regarding the possibility of earthquake prediction by divination, and indeed no influential Japanese works on earthquakes regarded them as amenable to divination in advance. The new idea that earthquake precursors exist and might be useful in predicting a main shock derived from yin-yang explanations of earthquakes. Starting in the nineteenth century, certain phenomena that suggested excess heat or emissions of hot vapor from within the earth came to be routinely listed as precursory phenomena. Examples include unseasonably warm weather, steam or hot mist, stars or the moon appearing red, peculiar clouds in the sky, changes in well water, emissions of light, and (especially after 1855) strange animal behavior. Significantly, contemporary earthquake prediction in Japan and some other parts of the world is based on the same basic claim that we find in Thoughts on Earthquakes and other nineteenth-century works: if only we can catalogue and study a sufficient variety of precursors, we might be able to predict earthquakes. Moreover, many of the specific precursory phenomena claimed by contemporary advocates of this approach are the same phenomena found in nineteenth-century texts. I examine this topic in detail in chapter 6. The purpose of this section is to survey the major varieties of early modern precursors and to take note of their theoretical basis, a point often ignored by modern and contemporary advocates of the value of these phenomena for earthquake prediction.
Light Flashes, Thunder, Lightning
Perhaps more than any phenomenon other than shaking, Japanese lore associated earthquakes with flashes of light. Often this association included thunder and lightning, which many early modern books regarded as essentially the same phenomenon as earthquakes but occurring above the ground. Nineteenth-century accounts of earthquakes usually included a report on weather conditions, and sure enough, clouds and thunder were usually present prior to the shaking.
Reports of light flashes seem to have begun in 1662. Asai Ryōi described “shining objects” (hikarimono) in the skies that traveled to the top of Mt. Hiei, which were witnessed by many during the earthquake. Other documents from the Kanbun earthquake mentioned shining objects, and the Konoe Diary (Konoe nikki) even included a small line drawing of such an object, resembling a spoon or ladle, or possibly a comet. From that time to the present day, flashes, objects, or pillars of light have become common elements in Japanese descriptions of earthquakes. Moreover, in modern times, “earthquake lights” have been widely reported throughout the globe. Despite many hypotheses about these phenomena, no one has been able fully to describe these lights, much less determine their cause. In early modern Japan, the likely reason for the association of earthquakes and light flashes was the emerging consensus about the mechanical cause of earthquakes, which regarded them as underground thunder caused by explosive upwelling of yang energy. Thunder, of course, was closely connected with lightning. In this way, earthquakes and rainstorms also became associated in the popular imagination. Foundation Stone was not the origin of the association of earthquakes and light flashes, but it helped to popularize the connection. During the Genroku earthquake, “flashes of light that resembled lightning” moving toward the southeast could be seen after the earthquake, and the earth emitted a roar, “like thunder” just before the earth shook. Similarly, 152 years later, the Ansei Edo earthquake began with a light show according to many accounts. “In the east, a flash of light appeared, similar to lightning. A moment later it was gone.” “As soon as a suspicious shining object flashed across every direction, the earth groaned.” In some versions, the light flashes emanated from or moved across the night sky: “In the southeast of Edo, a shining object resembling fire crossed the sky.” In other descriptions, the light came from the ground up: “At the time of the present earthquake, fire ki emanated from within the earth.” Workers on ships in the harbor reported seeing flashes of light over Edo at the time of the earthquake. Particularly spectacular is the account in Seeds of Tales from the Great Edo Earthquake of the Latter Age (Edo Ōjishin matsudai hanashi
no tane), an anonymous, sensational book from 1855. It tells of a series of light flashes, “white ki,” that emanated from fissures in the ground from the Nihon bank of Shin-Yoshiwara to Sensōji. This white ki struck the ninewheel spire (kurin) atop the temple's five-story pagoda, bending it before dispersing in all directions. Seeds of Tales from the Great Edo Earthquake of the Latter Age includes a dramatic image of a massive beam of light emerging from the ground at a forty-five-degree angle and causing an explosion above the spire. Fujiokaya Diary (Fujiokaya nikki) explained that a “shining object” traveling across the sky from north to south caused the spire to bend. The bent spire became an iconographic emblem of the earthquake, appearing in popular broadsides, catfish prints, journalistic literature such as Ansei Chronicle, and even in a newspaper article from the Meiji period. Dramatic emissions of light were reported widely in 1896 and 1933 in connection with the massive earthquakes and tsunamis off the Sanriku coast in northeast Japan. Meiji Sanriku earthquake witnesses also frequently reported booming cannonlike or thunderlike noises. Shōwa Sanriku earthquake investigators noted lightinglike flashes of light reported throughout the region. Although seismologists have never been able to identify plausible sources for light flashes or related phenomena, it was not for lack of trying. Writing in the 1950s, Musha Kinkichi, for example, defiantly urged students of earthquakes to take these descriptions of flashes of light seriously. Not long after Musha's time, the increasing acceptance of plate tectonics may have helped reduce interest in such phenomena, at least among seismologists. For example, Miki Haruo, in a late 1970s book on the Kyoto earthquake, gives only passing mention of light flashes and concludes that “the identity of the shining objects has been unknown from then to now.” Later, he concludes that there is no statistical relationship between earthquakes and rainstorms.
As we have seen, however, reports of and interest in earthquake lights continue to the present. The sheer volume of such reports is the main reason many seismologists will not dismiss the phenomenon outright. Earthquake lights, strange animal behavior, and other elusive possible precursors are typically identified in retrospect, after the earthquake has struck. I would argue that it is important to bear in mind the psychological dimension of these observations. Stated simply, people tend to see—or to believe they saw—what they expect to see. The early modern understanding of earthquakes as an underground version of thunder and lightning was especially conducive to perceiving flashes of light when the earth shook.
-  Asai, Kaname’ishi, 13–39. See also Kitahara, “Saigai to jōhō,” 236.
-  NJS, hoi (supplement), 149. The rounded, substantial part is labeled “front,” and the part resembling a handle or tail is labeled “back.” The text explains that this object emerged from a giant star, flew across the sky from the southwest (hitsujisaru) to the northeast (ushitora), and disappeared. Clearly, it was not seen as a comet, but a comet may have been the only frame of reference for conceiving of the object’s shape.
-  For a summary of earthquake lights, see Susan Hough, Predicting the Unpredictable: The Tumultuous Science of Earthquake Prediction (Princeton, NJ: Princeton University Press, 2010), 135.
-  “Kanrosō,” in DNJS, vol. 1 (kō), 287. See also Musha Kinkichi, Jishin namazu (Meiseki shoten, 1995, originally published in 1957), 53. Musha, who took this and other accounts of light flashes seriously as an integral component or concomitant of earthquakes, proposed that these light flashes were the result of a series of aftershocks.
-  Anonymous, Jishin narabi ni shikka saikenki (publisher unknown, 1855), 8–9 (text spanning the fifteenth and sixteenth page faces). See also Musha, Jishin namazu, 58.
-  “Jifūroku,” in DNJS, vol. 2 (otsu), 537.
-  Ōsaki Shōji, ed., Kinsei Nihon tenmon shiryō (Hara shobō, 1994), 408. I thank Laura Nenzi for this reference.
-  “Chika yori kaki o hassuru no jō,” in AKR, vol. 3, 11. See also Arakawa, Jitsuroku, Ō-Edo kaimetsu no hi, 88, and Musha, Jishin namazu, 58.
-  For example, “Ansei itsubō jishin kibun,” in NJS, vol. 5, supplement 2, part 1, 434.
-  Anonymous, Edo Ōjishin matsudai hanashi no tane, 1855, 11. See also Musha, Jishin namazu, 59, and Abe Yasunari, “Jishin to hitobito no sōzōryoku,” in Chūō bōsai kaigi, 1855 Ansei Edo jishin hōkokusho (Fuji sōgō kenkyūjo, 2004), 131.
-  FN, 516. For a complete listing of these and many other documents from this earthquake dealing with emissions of light and clouds, see SGS, vol. 2 (ge), 951–955.
-  AKS, vol. 2, illustration and text as insert between 10 and 11. See also Arakawa, Jitsuroku, Ō-Edo kaimetsu no hi, 152. The Kenmonshi account simply reports the damaged spire in the context of damage to other shrines and temples. For other examples of the damaged spire in popular prints and books, see Nakayama Einosuke, “Ansei no kyodai jishin to kawaraban,” in Inagaki Funio, ed., Edo no taihen: Jishin, kaminari, kaji, kaibutsu (Heibonsha, 1995), 76–77. For “catfish” print examples (without catfish), see #6 and #22 in Miyata Noboru and Takada Mamoru, eds., Namazue: shinsai to Nihon bunka (Ribun shuppan, 1995), 138, 241–242, 251–252. Print #6 does not mention the beam of ki or light but instead attributes the bending to deterioration of the spire’s core support, which had not been replaced in 136 years. For the Meiji example, see “Jishin to arashi,” in Yomiuri shinbun, January 4, 1885, special edition, 3.
-  Yoshimura Akira, Sanriku kaigan ōtsunami (Bungei shunjū, 2004), 82–87.
-  Musha, Jishin namazu, 52–104. Musha argues that accounts of flashes of light, pillars of fire, and so on “should not be carelessly dismissed as absurd. It is hardly the case that people of the past purposely wrote lies. . . . Accounts of things such as earthquake light could not have been written by the imagination” (52). He then goes on to analyze the matter for over fifty pages. Musha’s earliest example is from 1257, and the next one is from 1703. He seems not to have been aware of Foundation Stone and other Kanbun earthquake materials.
-  Miki Haruo, Kyōto daijishin (Shibunkaku shuppan, 1979), 60–66.
-  Drawing on the work of psychologist Kikuchi Satoshi, Shimamura explains that the common phenomenon of the discovery of precursors after the fact of an earthquake is an example of “erroneous correlation” (sakugo sōkan), whereby preconceived notions combined with the shock of a major earthquake can result in the recall, or manufacture, of precursors from events vaguely remembered. See Shimamura, Jishin no gimon, 102–103.